def build_rs_frequs(self, variables_calculation: list, filter_string: str=None, variables_calculation_avg: list=None, title: str=None,
codeplan: CodePlan=None) -> ReportingSet:
"""Build a reportingset based on frequencies.
:param variables_calculation:
:param filter_string:
:param variables_calculation_avg:
:param title:
:param codeplan:
"""
self.check_setup()
table_column_names = self.db.query_table_column_names(self.database_table)
for variable_name in variables_calculation:
if variable_name not in table_column_names:
raise Exception("Variable '{}' not found in {}!".format(variable_name, self.database_table))
self.variables_calculation_avg = variables_calculation_avg # XXX umbenennen
rs = ReportingSet(title=title)
rs['TOTAL'] = self._calc_frequencies(variables_calculation=variables_calculation, filter_string=filter_string)
if codeplan:
rs.split_main = codeplan
rs.split_main.key_order + [key for key in rs['TOTAL']['COUNT'] if key not in codeplan] # # Add missing codes
keys_sorted = None
else:
if len(rs['TOTAL']) > 0:
rs.split_main.data = list(rs['TOTAL']['COUNT'].keys())
keys_sorted = None
if self.order_freq_desc:
keys_sorted = sorted(copy.copy(rs['TOTAL']['COUNT']).items(), key=operator.itemgetter(1))
keys_sorted.reverse()
rs.split_main.key_order = [chunk[0] for chunk in keys_sorted]
if 'OTHER' in rs.split_main.key_order:
rs.split_main.key_order.remove('OTHER')
rs.split_main.key_order.append('OTHER')
rs.split_main.variables = variables_calculation
if self.result_subsplits:
self.calc_splits_sub(rs, filter_string, variables_calculation)
rs.timestamp = datetime.datetime.now().strftime('%d.%m.%Y %H:%M:%S')
return rs
def _build_table_tex(self, rs: ReportingSet, head=None):
"""
:param rs: ReportingSet to be texed
:param head:
:return:
"""
rs_txt = ""
codes_displayed_max_number = 90 # FIXME XXX not working.
mainsplit_width = self.split_main_width_cm
mainsplit_columns_count = 1
table_heads = []
data_columns_count = len(self.columns_reporting)
if not head:
table_heads.append(CodePlan()) # add empty codeplan if no head is present
elif isinstance(head, CodePlan):
data_columns_count *= len(head) + 1 # + 1 for total
table_heads.append(head)
elif isinstance(head, TabHead):
if head.is_subhead: # unschön FIXME
table_heads = head
data_columns_count = len(self.columns_reporting) * (sum([len(element) for element in table_heads]) + 1)
else: # unschön
raise Exception("WTF")
else:
raise Exception("Unkown type found as table head." + str(type(head)))
"""Filtern auf relevante Datensets, 'TOTAL' gehört immer dazu."""
datasets_names_relevant = [head_sub.variables[0] + "_" + str(code)
for head_sub in table_heads
for code in head_sub.key_order]
datasets_names_relevant.insert(0, 'TOTAL')
datasets = {name: rs[name] for name in datasets_names_relevant if name in rs}
# Bereinigen der Labels. Wenn mehr als 8 labels, dann nur die Labels mit Werten.
if len(rs.split_main.key_order) >= codes_displayed_max_number:
tempOrder = []
for key in rs.split_main.key_order:
if datasets['TOTAL']['COUNT'] and key in datasets['TOTAL']['COUNT'].keys():
tempOrder.append(key)
codes_trailing = (96, 99) # 'Weiß' nicht und 'keine Angabe' werden an das Ende gesetzt. TODO parametrisiere
for code in codes_trailing:
if code in tempOrder:
tempOrder.pop(tempOrder.index(code))
tempOrder.append(code)
rs.split_main.key_order = tempOrder
# Sonderschleife fuer den Titel. Der Titel wird nur einmal in das Inhaltsverzeichnis übernommen.
if len(table_heads) == 0:
raise Exception("Warum ist der Tablehead nicht >0?")
tableHead_tex = len(self.columns_reporting) * "&"
subHeadDataCols = 0
for head_sub in table_heads:
if len(head_sub) > 0:
tableHead_tex += r"&\multicolumn{" + str(len(self.columns_reporting) * len(head_sub)) + "}{l}{" + str(
self._convert_symbols2tex(head_sub.title)) + "}"
subHeadDataCols += len(head_sub)
tableHead_tex += r"\\" + "\n "
if data_columns_count - len(self.columns_reporting) > 0:
midrule_start = len(self.columns_reporting) + mainsplit_columns_count + 1
midrule_end = midrule_start + subHeadDataCols * len(self.columns_reporting) - 1
tableHead_tex += r"\cmidrule(l){" + str(midrule_start) + "-" + str(midrule_end) + "}"
#########################################
# ## Zeile mit Codes als Spaltennamen
#########################################
tableHead_tex += r"&\multicolumn{{{span}}}{{l}}{{GESAMT}}".format(span=len(self.columns_reporting))
codes_head = ['TOTAL'] + [head_sub.variables[0] + "_" + str(key)
for head_sub in table_heads
for key in head_sub.key_order]
# ## Kopfbeschriftung
if len(self.columns_reporting) < 3:
width = "1.9cm"
elif len(self.columns_reporting) == 3:
width = "4cm"
else:
width = "6cm"
for head_sub in table_heads: ####XXXXX warum fuer alle unterkoepfe?
for code in head_sub.key_order:
insert = self._convert_symbols2tex(head_sub[code])
if self._display_label == "both" and str(code) != insert:
insert = str(code) + " " + insert
tableHead_tex += r"&\multicolumn{{{columns}}}{{p{{{width}}}}}{{{insert}}}".format(
columns=str(len(self.columns_reporting)), width=width, insert=insert)
# ##XXXXXXXXXX brauche ich die rs vor dem STring??
tableHead_tex += r"\\ " + "\n "
##### Unterste Kopfzeile mit Spalteninhalt.
tableHead_tex += "Code Label" if self._display_label == "both" else "Kategorie"
tableHead_tex += ''.join(
["&\multicolumn{1}{c}{" + r"\footnotesize{" + self.column_type_labels[reporting_column_type] + "}}" for
reporting_column_type in
self.columns_reporting])
tableHead_tex += ''.join(
["&\multicolumn{1}{c}{" + r"\footnotesize{" + self.column_type_labels[reporting_column_type] + "}}" for code
in head_sub.key_order
for subHead in table_heads for reporting_column_type in self.columns_reporting])
# for subHead in tableHeads:
# for code in subHead.key_order:
# for column in self.reportingColumns:
# dummy += "&\multicolumn{1}{c}{"+r"\footnotesize{"+self.columnType[column]+"}}"
tableHead_tex += r"\\"
rs_txt += r"\begin{longtable}[l]{"
#### Spaltenbreite in Abhaengigkeit der dargestellten Codes.
rs_txt += "p{1cm}" if self._display_label == "code" else r">{{\raggedright}}p{{{width}cm}}".format(
width=mainsplit_width)
rs_txt += "r" * data_columns_count + "}\n"
# ## Ueberschrift, mit Pruefung, dass nur die erste Tabelle jeweils in Inhaltsverzeichnis kommt.
rs_txt += r"\caption"
if self.lagTitle == rs.title:
rs_txt += r"[]"
insert = None
if not rs.title:
rs.title = rs.split_main.variables[0] # falsche reihenfolge??? #XXXX waru mnicht alles;
elif self.debug_content:
insert = " [{}]".format(rs.split_main.variables[0]) # reihenfolge??? #XXXX waru mnicht alles;
pass
if not insert:
insert = ""
self.lagTitle = str(rs.title)
if 'section' in rs.content_misc.keys():
rs_txt += r"{" + self._convert_symbols2tex(rs.content_misc['section'] + " -- " + rs.title) + r"}\\ "
else:
if rs.title is None:
rs.title = ''
rs_txt += r"{" + self._convert_symbols2tex(rs.title + insert) + r"}\\ "
rs_txt += r"\toprule[1.5pt]\\ " + tableHead_tex + r"\midrule[1pt]\addlinespace \endfirsthead "
head_last = r"\caption[]{" + self._convert_symbols2tex(rs.title) + r" \dots (Fortsetzung)}\\ "
head_last += r"\toprule[1.5pt]\\ " + tableHead_tex + r"\midrule[1pt]\addlinespace \endhead "
foot_first = r"\bottomrule[1.5pt] "
foot_first += r"\multicolumn{" + str(data_columns_count + 1) + r"} {l} {\textit{Fortsetzung \ldots}}\\"
foot_first += r"\endfoot "
foot_last = r"\bottomrule[1.5pt] "
if rs.content_misc.get('subTitle'):
foot_last += r" \multicolumn{" + str(
data_columns_count + 1) + r"}{p{0.5\linewidth}}{\footnotesize{" + self._convert_symbols2tex(
rs.content_misc['subTitle']) + r" }} \\"
if rs.content_misc.get('comment'):
foot_last += r" \multicolumn{" + str(
data_columns_count + 1) + r"}{p{0.5\linewidth}}{\footnotesize{" + self._convert_symbols2tex(
rs.content_misc['comment']) + r" }} \\"
if rs.content_misc.get('filterText'): # TODO wieso get?
foot_last += r" \multicolumn{{{columnCount}}}{{p{{0.5\linewidth}}}}{{\footnotesize{{Filter: ".format(
columnCount=str(data_columns_count + 1)) + self._convert_symbols2tex(
rs.content_misc['filterText']) + r" }} \\"
if rs.timestamp:
foot_last += r" \multicolumn{" + str(
data_columns_count + 1) + r"}{p{0.5\linewidth}}{\footnotesize{" + \
self._convert_symbols2tex(rs.timestamp) + r" }} \\"
foot_last += r"\endlastfoot "
###############
############### Ab hier Daten
###############
data_block = ""
total = {column: {} for column in self.columns_reporting}
i = 0
for row in rs.split_main.key_order:
"""zeile fuer zeile Daten eintragen """
if i > self.split_main_codes_max:
break # XXX TODO: sdfasd
i = i + 1
if self._display_label == "both":
# dataBlock += str(row)+" "+self._convertSympols2Tex(rs.mainSplit.data[row])
try:
data_block += self._convert_symbols2tex(str(row)) + " " + self._convert_symbols2tex(
rs.split_main[row])
except KeyError:
data_block += self._convert_symbols2tex(str(row)) + " " + self._convert_symbols2tex(
rs.split_main.get(row, ''))
else:
# if isinstance(rs.mainSplit.data[row], int):
# if isinstance(rs.mainSplit[row], int):
data_block += rs.split_main[str(row)] if isinstance(rs.split_main[(row)],
int) else self._convert_symbols2tex(
rs.split_main[row])
# try:
# data_block += unicode(rs.split_main[str(row)]) if isinstance(rs.split_main[(row)], int) else self._convert_symbols2tex(
# rs.split_main[row])
# except KeyError: # ##XXXX iuebergang
# row = int(row)
# tempval = rs.split_main.get(row, '')
# data_block += unicode(tempval) if isinstance(tempval, int) else self._convert_symbols2tex(tempval)
# dataBlock += unicode(rs.mainSplit.data[row])
# else:
# dataBlock += self._convertSympols2Tex(rs.mainSplit[row])
# dataBlock += self._convertSympols2Tex(rs.mainSplit.data[row])
column_counter = 1 # # XXX kann ich das auch anders machen?
for code in codes_head:
for reporting_column_type in self.columns_reporting:
if not column_counter in total[reporting_column_type]:
total[reporting_column_type][column_counter] = 0
# # XXXXX der Abgleich mit str hier ist wohl nur aufgrund eines Fehlers notwendig. ???
row = str(row)
if code in datasets and reporting_column_type in datasets[code] and str(row) in datasets[code][
reporting_column_type]:
total[reporting_column_type][column_counter] += datasets[code][reporting_column_type][str(row)]
if reporting_column_type == 'COUNT':
data_block += "&" + str(
locale.format("%.0f", datasets[code][reporting_column_type][row], True))
else:
data_block += "&" + str(locale.format("%." + str(self.decimals) + "f",
round(datasets[code][reporting_column_type][row],
self.decimals),
grouping=True))
else:
total[reporting_column_type][column_counter] += 0
data_block += "&"
column_counter += 1
self.columns_reporting
data_block += "\\\\ \n \\addlinespace "
if self.dataSepLines_enabled:
data_block += r" \midrule[0.2pt]"
if self.dataSepLines_enabled:
data_block = data_block[0:-15]
block_summary_tex = r"\midrule\addlinespace "
block_summary_tex += "GESAMT"
column_counter = 1
for code in codes_head:
for reporting_column_type in self.columns_reporting:
try:
if reporting_column_type == 'COUNT':
block_summary_tex += r"&" + str(
locale.format("%.0f", total[reporting_column_type][column_counter], True))
elif reporting_column_type == 'PCT':
block_summary_tex += r"&" + str(locale.format("%." + str(self.decimals) + "f", round(
total[reporting_column_type][column_counter], self.decimals)))
else:
block_summary_tex += r"&" + str(locale.format("%." + str(self.decimals) + "f", round(
total[reporting_column_type][column_counter], self.decimals), True))
except:
block_summary_tex += r"&"
column_counter += 1
block_summary_tex += r"\\ \addlinespace "
if 'TOTAL' in datasets and datasets['TOTAL']:
if 'COUNT_TOTAL' in datasets['TOTAL']:
block_summary_tex += "Fallzahl"
for code in codes_head:
if self.columns_reporting == ['PCT_GN_W']:
datasets[code]['COUNT_TOTAL'] = datasets[code]['COUNT_GN_TOTAL']
datasets[code]['COUNT_W_TOTAL'] = datasets[code]['COUNT_GN_W_TOTAL']
if datasets[code]:
count_total = str(locale.format("%.0f", datasets[code]['COUNT_TOTAL'], True))
else:
count_total = str(0)
for reporting_column_type in self.columns_reporting:
if reporting_column_type == "COUNT":
block_summary_tex += r"&" + count_total
elif reporting_column_type == "COUNT_W":
try:
block_summary_tex += r"&" + str(locale.format("%." + str(self.decimals) + "f",
round(datasets[code]['COUNT_W_TOTAL'],
self.decimals), True))
except:
block_summary_tex += r"&"
else:
block_summary_tex += r"&"
# if len(self.reportingColumns) > 1:
# summaryBlock_tex += r"&\multicolumn{"+str(len(self.reportingColumns))+"}{l}{"
# if datasets[code].data:
# summaryBlock_tex += count_total
# summaryBlock_tex += str(datasets[code].data['COUNT_TOTAL']) #### nachbessern
# except:
# summaryBlock_tex += str(0)
# else:
# summaryBlock_tex += str(0)
# summaryBlock_tex += r"}"
# else:
# summaryBlock_tex += r"&"+count_total
block_summary_tex += r"\\ \addlinespace "
if 'AVG_TOTAL' in datasets['TOTAL']:
block_summary_tex += "Durchschnitt"
for code in codes_head:
if datasets[code]:
count_total = str(
locale.format("%." + str(self.decimals) + "f", datasets[code]['AVG_TOTAL'], True))
else:
count_total = str(0)
for reporting_column_type in self.columns_reporting:
if reporting_column_type == "COUNT":
block_summary_tex += r"&" + count_total
elif reporting_column_type == "COUNT_W":
try:
block_summary_tex += r"&" + str(locale.format("%." + str(self.decimals) + "f",
round(datasets[code]['COUNT_W_TOTAL'],
self.decimals), True))
except:
block_summary_tex += r"&"
else:
block_summary_tex += r"&"
# if len(self.reportingColumns) > 1:
# summaryBlock_tex += r"&\multicolumn{"+str(len(self.reportingColumns))+"}{l}{"
#
# if datasets[code].data:
# summaryBlock_tex += count_total
# summaryBlock_tex += str(datasets[code].data['COUNT_TOTAL']) #### nachbessern
# except:
# summaryBlock_tex += str(0)
# else:
# summaryBlock_tex += str(0)
# summaryBlock_tex += r"}"
# else:
# summaryBlock_tex += r"&"+count_total
block_summary_tex += r"\\ \addlinespace "
# if 'COUNT_W_TOTAL' in datasets['TOTAL'].data.keys():
# summaryBlock_tex += "Fallzahl gewichtet"
# for code in codes_head:
# if len(self.reportingColumns) > 1:
# summaryBlock_tex += r"&\multicolumn{"+str(len(self.reportingColumns))+"}{l}{"
# summaryBlock_tex += str(locale.format("%."+str(self.decimals)+"f", round(datasets[code].data['COUNT_W_TOTAL'], self.decimals)))
# summaryBlock_tex += r"}"
# else:
# summaryBlock_tex += r"&"+str(locale.format("%."+str(self.decimals)+"f", round(datasets[code].data['COUNT_W_TOTAL'], self.decimals)))
# #summaryBlock_tex += r"&0"
# summaryBlock_tex +=r"\\ \addlinespace "
if False:
aready_displayed = ["TOTAL", ]
block_summary_tex += "Zusammenhang"
for code in codes_head:
if code == "TOTAL":
block_summary_tex += r"&\multicolumn{" + str(len(self.columns_reporting)) + "}{l}{/}"
elif code[0:code.rfind('_')] not in aready_displayed:
aready_displayed.append(code[0:code.rfind('_')])
statistics = rs.subSplits[code[0:code.rfind('_')]].statistics
block_summary_tex += r"&\multicolumn{" + str(len(self.columns_reporting) * len(
rs.subSplits[code[0:code.rfind('_')]].key_order)) + "}{l}{"
block_summary_tex += "P($\chi2$): " + str(
round(statistics["chi2P"] * 100, 1)) + "\%, CramersV: " + str(
round(statistics["CramersV"], 2))
block_summary_tex += r"}"
else:
pass
block_summary_tex += r"\\ \addlinespace "
rs_txt += head_last + block_summary_tex + foot_first + block_summary_tex + foot_last + data_block
rs_txt += " \end{longtable}\n\r\\newpage\n\r"
return rs_txt
